Unbiased genome sequencing has recently uncovered recurrent mutations in several splicing factors across multiple neoplasms. SF3B1 is the most commonly mutated splicing factor, especially common in a clonal bone marrow disorder called myelodysplastic syndrome (MDS). SF3B1 mutations are single-allele, non-synonymous (each changing a single amino acid) and mutually exclusive, suggesting a gain-of-function physiology. The current, prevailing model for pathogenesis of splicing factor-mutant diseases assumes that aberrant splicing induces dysregulated expression of a few key genes that leads to clonal evolution and subsequent disease phenotypes. Comprehensive transcriptome analysis of patient samples and experimental models by our group and others reveal the use of novel or cryptic 3' splice sites (3'SS) in SF3B1 mutant cells. Our preliminary results also show that these cryptic 3'SS are normally not used since they are sequestered within secondary structures of nascent pre-mRNA; this constraint is overcome by SF3B1 mutant spliceosomes. In this proposal, we seek to resolve important knowledge gaps involving mutant SF3B1 biology and the mechanistic basis of associated disease physiology through innovative molecular approaches in distinct model systems. In Aim 1, we will definitively answer an important question pertinent to 3'SS selection: does cryptic 3'SS selection in SF3B1-mutant spliceosome result from a change in branch point (BP)? We will use CLIP-Seq (cross-linked immunoprecipitation and sequencing) to identify direct RNA targets of SF3B1 binding and Lariat-Seq (sequencing of intronic lariat intermediates) to determine altered BP choices in mutant SF3B1. In Aim 2, we will build on our preliminary results that show a direct interaction between SF3B1 and chromatin that is regulated in a cell-cycle dependent manner. We hypothesize that chromatin-dependent SF3B1 functions are disrupted in SF3B1-mutant cells resulting in dysregulated gene expression and/or RNA splicing. In Aim 3, we will define the physiologic consequences of specific gene expression abnormalities resulting from SF3B1 mutations in relationship to MDS biology. Our experiments are designed to elucidate how aberrant splicing and expression of particular transcripts perturb mitochondrial function and alter hematopoiesis in vivo. Importantly, we now recognize that molecular outcomes of SF3B1 mutations are dependent on cellular context and species-specific determinants, due to poor conservation of intronic domains that SF3B1 recognizes. Hence, we will use complementary models in human induced pluripotent stem cells and murine embryonic stem cells edited for CRISPR-Cas9 system for these studies. The application brings together two groups with complementary expertise in RNA biology (Pillai and Neugebauer) and hematopoiesis and cell cycle control (Minella). Defining precise mechanisms of splicing alterations in SF3B1-mutant disease will inform the development of novel therapies directed at mutant proteins themselves or dysregulated downstream pathways.